Vary Framework and Its Utility in Risk Management - Coursework Example

Number Vary framework and its Utility in Risk Management INTRODUCTION Risk happens to be unavoidable. It is the key ingredient for profit generation within whatever market sensitive activity. An investor’s viewpoint of risk is it is all about losing money. VaR is founded on the same common since fact. Through the assumption that venture capitalists care about odds of really huge losses, VaR seeks to answer the question, what could be one’s worst case scenario.” or even “how much would one loose within a really bad period.” VaR happens to be some single, summary, statistical measure for possible losses of portfolio. It indicates the highest possible loss amount, which some portfolio will likely lose within a given time period at specified confidence level. A good example would be 95% daily VaR for $1 Million, could mean the likely hood for the same portfolio to lose over a million dollars within a worst day happens to be below 5%. In no way does this mean that such portfolio may not lose over a million dollars. The truth is that over one hundred days, the portfolio would be expected to lose over $1 million for five time approximately. In addition, this does not mean that an individual would not collectively loose significantly more along a longer horizon. Banks, mutual funds, hedge funds as well as other financial service companies or even brokers can utilize value at Risk. Most of such firms use VaR in prediction of size of outlying losses of the future, or even gains that their portfolios or those of their clients might experience (Ran & Jin, 2008: p 1). Most firms make use of VaR in the determination of needed collateral from an execution customer for some margin loan utilized in trading financial instruments, for instance. Buy-side entities like hedge funds make use of VaR in determining whether the allocation of a portfolio does exceed investment mandate or a current risk tolerance (BPL, 2015, 2). BACKGROUND Despite the fact that VaR was not used broadly before mid-1990s, the measure’s origin date further back in time. Markowitz Harry and others developed the mathematics, which underlie VaR in portfolio theory context (Glyn, 200:p 32). However, their efforts aimed at a different destination (devising equity investors’ optimal portfolios). Specifically market risk focus as well as the co-movements effects in such risks are core to the manner in which VaR is computed (Ronald, Kees, & Rachel, 1999: P 2). The Motivation for VaR measures utilization, though, arose from the crises, which affected financial service firma through time as well as supervisory responses to such disasters. The 1st supervisory capital requirements for banks got indorsed within the outcome of the great depression as well as the era’s bank failures, at the time when Securities Exchange Act introduced the SEC OR Securities Exchange Commission, then demand that banks maintain their borrowings below two thousand percent of the equity capital of the banks. With increased risk brought about by the arrival of derivative markets as well as floating exchange rates within early 1970s, there was refinement as well as expansion of capital requirements within the Uniform Net Capital Rule of SEC, which was promulgated in the year 1975 (Split History, 2015). It categorized financial assets held by banks into 12 classes, based on risk, as well as required distinct capital requirements for each one of them, ranging from 0 percent for the short-term treasuries – 30 percent for equities. The banks were demanded to give reports on capital calculations within quarterly statements, which were tattled FOCUS (Financial and Operating Combined Single) reports. The 1st regulatory measures, which induced VaR got initiated in 1980, the time when SEC tied capital requirements for financial service companies to losses to be incurred, a 5 percent confidence across a 30-day interval, within distinct security classes (such potential losses were computed by use of historical returns). Despite that, the measures could be described as haircuts but not as capital or Value at risk, SEC was demanding that financial service companies embark on a procedure of estimating 1 month 95 percent Value at Risk, as well as hold sufficient capital to cover possible losses. At the same period, commercial banks as well as investment trading portfolios were getting to be larger as well as more instable, hence there was need for a more classy as well as appropriate measures for risk control. In the year 1986, Ken Garbade in Bankers Trust, within internal documents, introduced Vary sophisticated measures for the fixed income portfolio of the firm based on covariance in different maturity bonds’ yields. Come the early 1990s, most financial service companies had established VaR basic measures; however, there were great variations in the manner in which it was measured (Glyn, 2002). As a result of many catastrophic losses relative to utilization of leverage and derivatives from 193 to 1995, concluding with Barings failure, the British Investment Bank, resulting from empowered trading within Nikkei futures as well as options by one Nick Leeson, an established trader within Singapore, companies became ready for more inclusive risk measures. Come 1995, J.P. Morgan allowed public to access data on core variances across and variances on different security as well as asset categories, which it internally utilized for close to a decade in managing risk. The service was titled “RiskMetrics.”Value at Risk was used for describing the risk measure, which developed from the data. This measure found some ready audience with investment as well as commercial banks as well as supervisory authorities overseeing them. Within the last decade, Value at Risk has come to be the established measure for risk exposure within financial service companies while it has started finding acceptance within different firms (Kyriakos, 2005). ANALYSIS VaR may be estimated through different methodologies. In general, they rely upon computer algorithms, which use statistical models for quantifying portfolio risk. The key ideas behind these approaches is constructing the distribution for all possible values of portfolio over the given period then infer VaR from the distribution in accordance with the specified confidence level. Variance-Covariance Approach Since VaR measures probability that a portfolio of asset’s value is going to drop below a given value within specific period, then it should be comparatively simple to calculate if it is possible to derive some probability distribution for probable values. This is what happens basically within the variance-covariance approach, which is an advantageous method in term of simplicity. However, it is limited with difficulties related to deriving probability distributions. The approach involves 4 main steps: Step 1 requires that each asset within a portfolio is taken and mapped on to more simpler standardized instruments. Example, a 10-year old, which has annual coupon B, could be broken to 10, 0-coupon bonds with cash flows that match. The 1st coupon matches to a 1-year 0-coupon bond, which has a face value of B’ 2nd coupon with a 2-year 0-coupon which has face value of B. This goes similarly till the 10th cash flow that gets matched up with ten-year 0-coupon bond, face value of FV plus B. for much complex assets, mapping procedure happens to be more complicated, however, rudimentary perception does not change (SAP Company, 2015). In its place, estimating variances as well as covariances of numerous individual assets, estimation is done for statistics for mutual market risk instruments, which such assets get exposed to. Within step 2, every asset gets specified as a set for positions within standardized market instruments. It happens to be simpler for a ten-year coupon bond, when the intermediate 0-coupon bonds take face values, which match the coupons as well as the final 0-coupon bond take the face value adding he coupon within the period. With mapping, this procedure happens to be more complex when dealing with convertible bonds, derivatives, or stocks. After standardized instruments, which affect assets or asset within a portfolio has been identified, the variances within each of such instruments as well as covariances across such instruments have to be estimated within the next stage. Practically, such variances as well as covariance approximations are achieved through observing historical data. In the 4th step, VaR for portfolio is calculated by use of weights on standardized instruments that have been determined within the2nd step, as well as variances and covariances that were determined in step 3. Historical Simulation Approach This happens to be the cheapest means to estimate VaR for many portfolios. A portfolio’s VaR is estimated through forming a hypothetical time series for the portfolio’s returns, which is achieved through running such portfolio through real historical data as well as computing changes, which could have happened within each period (FARID, 2011). In running historical simulation, the first thing is the time series data for every market risk factor like in the variance covariance method. The data is not used for estimation of variances as well as covariances looking forward, due to the fact that changes within the portfolio with time produce all required information for calculating VaR. in the case of distributional assumptions, this method is agnostic while determination of Vary is through actual price movements (no assumptions underlying that normally drive the conclusion). Every day within time series bares an equal weight with regard to VaR measuring, a possible problem where there is trend within the variability. This approach is founded on an assumption that history will repeat itself. Here, the period in use provides complete and full snapshot of risks the market may be exposed to during other periods (Kyriakos, 2005: p 8). Monte Carlo Simulations Monte Carlo Simulations are quite useful in VaR assessment, with focus on probabilities for losses beyond a given value rather than the total distribution. Steps one and two in this approach reflect steps one and two in variance- covariance approach where market risks influencing assets of asset within a portfolio are identified, then individual assets are converted into positions within standardized instruments. Differences emerge within step three. Here, a simulation route is taken where probability distributions for every market risk factor are identified and the manner in which the factors move together is identified. While estimation of limits happens to be cheaper under normal distribution assumption for all variables, Monte Carlo simulations’ power comes from freedom to pick alternative distributions for variables. Subjective judgement can be brought in to modify such distributions (Palisade Corporation, 2015). After specification of the distributions, the process of simulation begins. Within every run, market risk variables yield different outcomes while the portfolio value reflects outcomes. Upon a repeated runs’ series numbering, normally in thousands, a portfolio values’ distribution is attained: one that may be utilized in assessing VaR. Comparing approaches All the three methods have their individual advantages as well as come with baggage. Variance-covariance method with its delta gamma and delta normal variations require strong assumptions to be made about return distributions for standardized assets, however, it happens to be cheap to compute after the assumptions are made. Historical simulation approach does not require any assumptions on return distributions nature, however, indirectly assumes that data utilized within simulation happens to be representative sample for the risks looking forward. Monte Carlo Simulations method facilitates the highest flexibility in choosing returns distributions as well as incorporating subjective judgement s as well as external data, on the other hand, the most demanding as far as computations are concerned (Amit, Max, Sonja, & Poppensieker, 2012: p 3). The answers that are gotten from the three methods happen to be a function of inputs. The variance-covariance approach and the historical simulation approach are going to render a similar result of VaR where historical returns data happens to be distributed normally and is utilized in estimating variance- covariance matrix. Likewise, Monte Carlo simulations and variance covariance approach will result to almost similar vales given that all inputs within the latter get assumed to be distributed normally with consistent variances and means. The answers will as well diverge as assumptions diverge. Monte Carlo and historical simulations methods will converge where distributions utilized in the latter happen to be in total based on historical data. What approach will yield the correct VaR is a question that a person may ask. The answer to this would be dependent on both the risks under assessment as well as the manner in which competing approaches are utilized. There are variants, which have developed in each of the approaches with the aim to improve performance. Most comparisons across methods happen to be skewed because researchers who do such comparisons test variants of an approach, which they developed against alternatives. Unsurprisingly, they find out that their methods even work better compared to the alternatives. It is through looking at a task at hand that the best VaR method can be identified, where one is assessing VaR for portfolios, which does not include any options, over a short time, the best approach will be the variance –covariance approach, nonetheless, its heroic normality assumption. If VaR is computed for risk source, which is stable as well as where substantial historical data exists, then historical simulations are the best approach here. Within most universal case of determining VaR for nonlinear portfolios over longer periods, within which historical data happens to be volatile as well as non-stationery and normality assumption happens to be questionable, the best approach here is Monte Carlo Simulations. CASE STUDIES Case Study 1.VaR Masked JP Morgan two billion dollar loss In May 2012, JP Morgan executive self-confessed that the company would likely go through volatility within its earnings during quarters ahead, as the company attempts to manage so systematic credit portfolio designed by the executive to hedge stressed environment credit, however, turned to be quite expensive for the firm. Dimon, the CEO exposed that the CIO’s VaR had approximately doubled from average utilization of $67 million in the 1st quarter to $129 million, upon scrapping the new model of the CIO’s and appropriately revising the figures. The company made a decision to return to the CIO used methodology in calculating VaR in 2011. In the first quarter, the company implemented some new VaR model that deemed adequate, and then went back for an old one utilized before that seemed more adequate. In 2012, VaR swung approximately within the range of $85 million to $187 million in the 1st quarter under this old model. In the year 2007 and 2008, Value at Risk came in for criticized while most models failed in predicting possible losses overwhelming most large banks ( pedrogor, 2015). The relationship intensely demonstrates potential for internal risk models for banks to produce massively different results that could have true economic influence. This case could as well bring into attention again the manner in which VaR model for banks could hinder and aid risk management (CHRISTOPHER, 2012). Case Study 2: Transforming global bank’s Approach to Market Risk Global investment bank was faring comparatively well at the time as well as after 2008 economic crisis. This was due to the bank’s decisive action of management. However, the disaster had put the market risk models of the bank at to test, exposing considerable issues with their projections accuracy. A 3-year program was lounged in response, to transform market risk tools of the bank as well as transform infrastructure focusing on reforming the VaR mode, economic capital model, as well as stress testing frameworks (McKinsey & Company, 2015). Discovery The VaR model overhaul involved some client major shift: from the Monte Carlo Simulations approach as well as sensitivity analysis based upon data residing within risk function’s system to the historical simulations approach as well as full revaluation based upon data existing in front office systems (common model in most banks). The project was structured trough creation of 4 working groups, which included market data, methodology, IT implementation, and procedure. The prove concept was an earlier pilot used to test the model by the team. To face-lift, the stress testing approach of the bank the team put in place some framework, which defined a broad range of shocks, which influenced economy. For creation of a stress test scenario that is customized, client managers mix as well as match such shocks at various levels of severity through some user interface. Previous, setting up as well as running stress-test scenario was a mutual procedure, which lasted 2 weeks. This new framework made it possible for the client to create stress test scenario within a day, hence using stress test scenario as the tool for managing day-to-day risk. An operating model was established while IT system was implemented for supporting the new tools. This ensured that the bank utilized the same systems while determining risk measures through the whole portfolio, something that led to a reliable market risk outlook. Impact The bank utilized new historical simulation approach for preliminary VaR estimates within some successful pilot program that involved a portfolio for interest rate products. This new model reduced complexity in process and system. Additionally, it made it possible for risk and front office managers to assess incremental effect for individual trades, enabling optimal hedging strategies selection. Pending regulatory support for the new approach, in two years, the firm plans to be utilizing historical simulations in the assessment of all market risk elements for the whole portfolio (McKinsey & Company, 2015). CONCLUSION VaR has advanced a risk assessment tool for banks as well as other financial service companies. Is utilization within such firms is driven by failure in systems for tracking risk up to the 1990s, to detect unsafe risk takings by traders and it provided a major advantage: a measure for capital risk during extreme conditions within trading portfolios, which might be adapted regularly. Banks and financial service companies may use conventional VaR modified by Hallerback and Menkveld, to accommodate numerous market factors. This measure may be utilized not only in determining the source of risk, but as well mage it in so better way, aiming at maximizing shareholder wealth. Conditional value of VaR, which is defined as the weighted average for VaR, could be well thought out as an upper bound on VaR and could minimize problems related to managers’ excessive risk taking. References pedrogor. (2015). Hedge Funds, Systemic Risk, And the Financial Crisis of 2007-2008. Retrieved from scribd.com: http://www.scribd.com/doc/17363186/Hedge-Funds-Systemic-Risk-And-the-Financial-Crisis-of-2007-2008#scribd Amit, M., Max, N., Sonja, P., & Poppensieker, T. (2012). Manging Market Risk Taday and Tomorrow. Markinsky Working Papers at Risk, 23. BPL. (2015). Introduction to Value at Risk (VaR). UNDERSTANDING MARKET, CREDIT, AND OPERATIONAL RISK, 20. CHRISTOPHER, W. (2012). Value-at-Risk model masked JP Morgan $2 bln loss. Retrieved from reuters.com: http://www.reuters.com/article/2012/05/11/jpmorgan-var-idUSL1E8GBKS920120511 FARID, J. (2011). Value at Risk Historical Simulation approach in Excel. Retrieved from financetrainingcourse.com: http://financetrainingcourse.com/education/2011/05/value-at-risk-histograms-and-risk-management-in-excel/ Glyn, A. H. (2002). History of Vlaue at Risk- 1992- 1998. 27. Kyriakos, A. (2005). Introduction to Value-at-Risk (VaR. Freight Metrics, 29. McKinsey & Company. (2015). Transforming a global bank’s approach to market risk. Retrieved from mckinsey.com: http://www.mckinsey.com/client_service/risk/case_studies/transforming_a_global_banks_approach_to_market_risk Palisade Corporation. (2015 ). Monte Carlo Simulation. Retrieved from .palisade.com: http://www.palisade.com/risk/monte_carlo_simulation.asp Ran, S., & Jin, Z. (2008). Value-at-Risk Based Portfolio Management in Electric Power Sector. 5. Ronald, H., Kees, G. K., & Rachel, A. P. ( 1999). Asset Allocation in a Value-at-Risk Framework. 26. SAP Company. (2015). Variance/Covariance Approach: Theoretical Background. Retrieved from sap.com: http://help.sap.com/saphelp_erp60_sp/helpdata/en/aa/a945aafa0611d1a5710000e839c3d0/content.htm Split History. (2015). VaR Split History. Retrieved from splithistory.com: https://www.splithistory.com/var/ Read More